The term “long/short” (LS) has wide and varied applications in the world of investments. In its broadest definition, an LS strategy signifies an investment process not constrained to holding only long positions. In perhaps its most common usage, an LS fund tends to refer to an equity hedge fund that takes both long and short positions. Whatever the application of the term, LS investing usually implies a search for absolute returns and an attempt to decrease or neutralize market beta. This combination of absolute returns with low market beta can also be described as “alpha” and has historically been considered an “active” investment approach. Here we examine the LS investing landscape from the perspective of an indexer. We make the case that there are sensible and investable ways to index these strategies, thereby creating a “passive” approach to various sectors of LS investing.
Certain questions arise in such an undertaking: Which types of strategies are suitable for indexing and which are not? What are the pitfalls of index construction when considering long/short strategies? When looking at LS fund results, how can we separate those attributed specifically to the superior skill of some managers from those managers who are able to deliver purely because of the unconstrained nature of the LS investing techniques at their disposal? To address these and other questions, we first provide an overview of the four main types of LS models. Next, we examine the case for passive approaches to LS investing. Finally, we present case studies of the Credit Suisse long/short index as an example of factor-based replication and the Credit Suisse merger arbitrage index as an example of mechanical replication.
The Long/Short Landscape
Whenever a manager decides to pursue a long/short strategy, her investment approach can be distilled into one of four models by answering certain questions about the approach: Are positions taken based on company- or issue-specific information? Are decisions primarily based on the outputs of a quantitative model or do they come from fundamental research? Is there a high or low degree of turnover?
There are four basic models used for selecting long/short positions in these strategies:
Valuation-based approaches are typically buy-and-hold strategies. The manager selects undervalued securities for the long side of the portfolio, and overvalued securities for the short side. Generally this approach leads to more stable positions with lower turnover than other strategies. Alpha in this space is derived from fundamental analysis of securities. Managers have a high amount of discretion in the selection process and focus on company fundamentals more than on the macro environment. Typically, these managers are long biased and, in addition to potential alpha from security selection, add value by varying their market exposure to deliver beta in a generally more efficient or better risk-adjusted format.
Trend-following models use technical indicators to identify trades. Typically, managers in this space use quantitative or systematic processes to generate buy or sell signals and have somewhat limited discretion in these decisions. This is a market-timing approach and is suitable across asset classes so long as the instruments traded are liquid. In addition, managers prefer instruments that are not company specific so as to remove idiosyncratic risk from the model. Returns from trend-following strategies are typically relatively uncorrelated to traditional beta, providing investors with diversification as well as the potential for alpha from the superior market timing of some managers’ models. A “managed futures” strategy would be an example of this approach.